Social Causality Prize I

[jabowery]

For those serious about showing the social pseudosciences to the exit of, well, "society", may I offer the Social Causality Prize I, now open at the X-Prize Foundation's heroX site.

The prize presents a wide range of social data and challenges those who believe they have the best theory of US society, to demonstrate its superiority via Ockham's Razor:

Produce the smallest program that outputs the provided social data file.

This test has a rigorous foundation in the theory of artificial general intelligence as accepted by founders of Google's DeepMind such as Shane Legg, who wrote he thesis on compression as the general intelligence benchmark.

[Garrett]

I'm sure you've thought of these things, so maybe you can explain a few questions that seem to pop out of the mental chute right away.

Wouldn't things like natural geographical resources, weather patterns and such things have either a direct or chaotic effect on census data? And wouldn't databases related to those be at least as large as the census data? If so, wouldn't that put a limit on compression using social causation theory?

[jabowery]

The theory upon which the Social Causality Prize I is based isn't "social causation theory" but algorithmic information theory.  Algorithmic information theory is the only formalization of evidence-based argument.  Molyneux should drop what he's doing and familiarize himself with it.